ANTI-AI ARCHIVE
ART-HISTORY NODE / 188 · 2022-08-02

Prompt-to-Prompt Image Editing with Cross Attention Control

Prompt-to-Prompt Image Editing with Cross Attention Control

Hertz, Amir; Mokady, Ron; Tenenbaum, Jay; Aberman, Kfir; Pritch, Yael; Cohen-Or, Daniel

Introduction

Prompt-to-Prompt studies cross-attention between words and image layout, using prompt changes for local replacement and broader transformation. It connects image synthesis with the problem of editable continuity.

ORIGINAL DOCUMENT#188
Figure 1: generated-image comparisons for word replacement, description strength and style changes.View full image ↗

Figure 1: generated-image comparisons for word replacement, description strength and style changes. Source: v1; the image version is distinct from the initial submission date.

Hertz, Amir; Mokady, Ron; Tenenbaum, Jay; Aberman, Kfir; Pritch, Yael; Cohen-Or, Daniel · original paper / arXiv · Rights in the paper and depicted works remain with their holders. Research quotation does not establish an open licence; republication rights await independent review.

Source · Prompt-to-Prompt Image Editing with Cross Attention Control ↗

RESEARCH ACCOUNT

Prompt-to-Prompt studies cross-attention between words and image layout, using prompt changes for local replacement and broader transformation. It connects image synthesis with the problem of editable continuity.

ANTI-AI ARCHIVE · Revised 2026-10-03

What the paper investigates

Simply changing a prompt can replace the whole image. This paper uses cross-attention as a control point, preserving or adjusting spatial relationships while words change. Demonstrations include word replacement, added descriptions and control over the extent to which a word affects the image.

Section sources: Prompt-to-Prompt Image Editing with Cross Attention Control

Its place in generative-art history

Editorial interpretation: the prompt becomes a means of revising an existing composition as well as initiating a sample. Reading this alongside real-image inversion makes continuity, the object of editing and the maker’s choices explicit historical questions.

Section sources: Prompt-to-Prompt Image Editing with Cross Attention Control

Reading limits and versions

Editing generated images does not establish reliable editing of arbitrary photographs; real images introduce an additional inversion problem.

Section sources: Prompt-to-Prompt Image Editing with Cross Attention Control

Sources for this account

Prompt-to-Prompt Image Editing with Cross Attention Control ↗

Read the arXiv abstract, authors and version history, plus the selected figure/page on PDF page 2; proofs and full experiments were not independently audited.

Read the arXiv abstract, authors and version history, plus the selected figure/page on PDF page 2; proofs and full experiments were not independently audited. The account and translation are AI-assisted, pending independent human review. Section references identify evidence without claiming independent verification of every historical statement.

Continue with a comparative question

Null-text Inversion for Editing Real Images using Guided Diffusion Models ↗

Prompt-to-Prompt: compare methods, control or evaluation conditions with Null-text Inversion.

GLIGEN: Open-Set Grounded Text-to-Image Generation ↗

Prompt-to-Prompt: compare methods, control or evaluation conditions with GLIGEN.

InstructPix2Pix: Learning to Follow Image Editing Instructions ↗

Prompt-to-Prompt: compare methods, control or evaluation conditions with InstructPix2Pix.

Adding Conditional Control to Text-to-Image Diffusion Models ↗

Prompt-to-Prompt: compare methods, control or evaluation conditions with Spatial control of generation.

Dates & version record

Date displayed for this node: 2022-08-02 · Historical date recorded for the source: 2022-08-02

The timeline uses the initial arXiv submission, distinct from conference publication, model release and the archive addition on 3 October 2026. The account and image refer to v1; its date is recorded in the source history.

These dates refer to the historical event or recorded version, not this page’s publication date. The original date precision and unresolved questions are retained.

Original sources & further reading

These links lead to the cited paper, article, institution or conference page. External texts retain their source languages.

01
Prompt-to-Prompt Image Editing with Cross Attention Control ↗

Prompt-to-Prompt Image Editing with Cross Attention Control

https://arxiv.org/abs/2208.01626

Current cited URL

Read the arXiv abstract, authors and version history, plus the selected figure/page on PDF page 2; proofs and full experiments were not independently audited.

Provenance, translation & verification

Archive node #188 · Initial research-paper submission; methods, evaluation and production conditions

Research materials & supplement references · 1

Recent research papers: additions and reconciled sources since 2022 · Read the arXiv abstract, authors and version history, plus the selected figure/page on PDF page 2; proofs and full experiments were not independently audited. · 004
Date as recorded: 2022-08-02
Prompt-to-Prompt Image Editing with Cross Attention Control ↗
Prompt-to-Prompt Image Editing with Cross Attention Control

Archive account based on the listed research materials, not a full translation of the linked work. AI-assisted translation; independent human review pending.

Official material was read within the stated scope; see the source note for reading limits and outstanding checks.

Cite this node

ANTI-AI ARCHIVE. Prompt-to-Prompt Image Editing with Cross Attention Control. Art-history node #188. https://salondesrefuses.cn/en/art-history/327

For specific historical claims, also cite the original sources above and include your access date. This account is not a full translation of the linked work.

Adjacent nodes follow chronological order; adjacency does not establish direct influence or causation.